AVR lost it's brains !

Go To Last Post
4 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think this was discussed a looong time ago, but I don't remember if there was any conclusive results, so, I bring it up again.

I just got two failing boards back from a customer. The CPU is the mega161-4AC.
On analysis, it turned out that both Flash memories were completely empty, all FF's.
The CPU's seemed otherwise okay, and after reprogramming, they worked nicely again.

No lockbits are set in the running systems, as they support firmware upgrade. The bootloader has a block, so it cannot reprogram itself.

The systems they're used in are pretty noisy, but the boards are power supplied from the USB port of a larger system. Also, a local regulator takes the voltage down to 3.3V, so power should be good. Anyway, no matter how much power drops there may be, the CPU should never loose it's Flash content.

This has been reported to Atmel, but I haven't got an answer yet.

In the meantime, has anyone else seen this happen ?

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello,

A long time ago I had that happen to me - once. AT90S8515 on a small robot was dead - no amount of reset or anything would help it. A quick program though and all was well again.

So I wouldn't say you've gone crazy ;-) Never had it happen again (that I remember), so I don't know what else to say...

-Colin

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not on the '161, but on older chips like the '1200 and '2313.

Out of curiousity, was the brown-out protection enabled on the chip that erased itself?

Also, were all I/O protected from high voltage spikes? The noisey environment might have triggered some sort of internal problem and thrown the chip into an erase mode. I had some problems a long time agon with Motorola MC68705P3 (UV microcontrollers) having their data direction bits change on the I/O ports from high frequency noise on the I/O. Perhaps something similar happned in your case.

Rick

--
"Why am I so soft in the middle when the rest of my life is so hard?"
-Paul Simon

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Brown out protection was not enabled, as this was buggy in the mega161.
I had an external DS1818 reset chip instead.

There is no direct I/O coming out, so that should not be the problem.

Anyway, there should be absolutely NO way for the chip to erase itself, no matter how noisy the I/O lines are.

/Jesper
http://www.yampp.com
The quick black AVR jumped over the lazy PIC.
What boots up, must come down.